Complement Fixation, Indirect Fluorescent-Antibody, Indirect ... comparison of cytomegalovirus antibody titers was made between the complement fixation test, a.
JOURNAL OF CLINICAL MICROBIOLOGY, Feb. 1984, p. 147-152 0095-1137/84/020147-06$02.00/0 Copyright © 1984, American Society for Microbiology
Vol. 19, No. 2
Immunity to Human Cytomegalovirus Measured and Compared by Complement Fixation, Indirect Fluorescent-Antibody, Indirect Hemagglutination, and Enzyme-Linked Immunosorbent Assays Department
JAMES A. BRANDT,'* JAMES D. KETTERING,2 AND JOHN E. LEWIS"2 Loma Linda University,2 and Department of Pathology and Clinical Medicine,
of Microbiology,
Loma Linda
University Medical Center,1 Loma Linda, California 92350 Received 21 June 1983/Accepted 24 October 1983
The complement fixation test is currently the test employed most frequently to determine the presence of antibody to human cytomegalovirus. Several other techniques have been adapted for this purpose. A comparison of cytomegalovirus antibody titers was made between the complement fixation test, a commercially available enzyme-linked immunosorbent assay, an indirect immunofluorescent technique, and a modified indirect hemagglutination test. Forty-three serum samples were tested for antibodies by each of the above procedures. The enzyme-linked immunosorbent, immunofluorescent, and indirect hemagglutination assays were in close agreement on all samples tested; the titers obtained with these methods were all equal to or greater than the complement fixation titer for 38 of the 41 samples (92.6%). Two samples were anticomplementary in the complement fixation test but gave readable results in the other tests. The complement fixation test was the least sensitive of the procedures examined. The commercial enzymelinked immunosorbent assay system was the most practical method and offered the highest degree of sensitivity in detecting antibodies to cytomegalovirus.
The possibility of transmitting cytomegalovirus (CMV) via blood donors with latent CMV infection was suggested as early as 1951 (39). An infectious mononucleosis-like illness identified as a postperfusion syndrome was shown to be caused by CMV (18, 19), and convincing evidence for blood as a source of CMV has been reported since that time (12). Recent studies have shown that up to 25% of exchangetransfused newborns may develop CMV infections (2). In one study (41), 13.5% of infants of CMV-seronegative mothers developed infection when exposed to at least one donor with serological evidence of latent CMV infection. Fatal or serious symptoms developed in 50% of the infected infants who had received more than 50 ml of packed erythrocytes. The use of CMV-seronegative donors virtually eliminated transfusion-acquired CMV infections in this group of subjects. CMV appears to be associated with the leukocyte-rich fractions of fresh whole blood (16, 21, 38). Evidence also indicates that the granulocyte fraction is more closely associated with CMV than is the mononuclear fraction (16). Complicating the picture is the fact that CMV has been recovered from the erythrocyte, plasma, and serum fractions of blood as well as from cerebrospinal fluid (1, 35). The solutions offered to the problem of transfusion-acquired CMV infection rely primarily on the use of leukocyte-poor blood or the use of donors who have no serological evidence of antibodies to CMV. Of these approaches to the problem, the use of CMV-seronegative blood appears to be the most practical. The intent of this study was to evaluate and compare several current methods for detecting the presence of antibody to CMV. The main purpose was to determine the most overall practical method for clinical laboratories to screen blood used to transfuse certain high-risk patients. The complement fixation (CF) test was chosen as the standard for comparison because of its long-standing and widespread use. *
The indirect fluorescent-antibody (IFA) technique, a modified indirect hemagglutination (IHA) test, and a commercially available enzyme-linked immunosorbent assay (ELISA) were chosen as examples of newer techniques that are currently in use and may offer certain advantages over the CF test.
MATERIALS AND METHODS Sera selection. The sera used were selected from samples submitted to the diagnostic virology laboratory at Loma Linda University Medical Center for determination of antibody titers to CMV. All sera had been previously tested for CMV antibody by the CF method and were stored frozen at -20°C. Sera were selected to include samples showing no anti-CMV activity, as well as samples representing low, medium, and high titers to CMV. CF assay. CF tests were performed with a standard microtiter method (31). A commercial antigen prepared from CMV strain AD-169 (lot no. 33015; Flow Laboratories, Inglewood, Calif.) was used. Obtained from the same manufacturer were guinea pig complement (lot no. 31009), hemolysin (lot no. 4788), CMV antigen control (lot no. 33471), and sheep erythrocytes (lot no. A4892). Serum titers were expressed as the reciprocal of the highest dilution in which 70% or more of the sheep erythrocytes were not hemolyzed. ELISA. Detection of antibody by the CMV Bio-Bead Titration Kit (Litton Bionetics, Kensington, Md.) is based on the principle of solid-phase immunoassay. A magnetic transfer device, supplied by the manufacturer, was used to transfer the solid phase from one reaction mixture to another. All reagents, controls, and equipment for performing the test were supplied in the kit. Ferrous metallic beads with bound CMV antigen from infected cell cultures were used in the test. Negative control beads had bound antigen prepared from noninfected cell cultures and were run in parallel with specific-antigen beads. The antibody titer was determined as the highest dilution of serum in which a 1 + to 2+ green color developed with the
Corresponding author. 147
148
J. CLIN. MICROBIOL.
BRANDT, KETTERING, AND LEWIS
antigen-coated bead. The degree of color development was compared with that yielded by the positive control sera at the stated titers. For the purposes of this study, the procedure outlined by the manufacturer was followed, with the single exception that the initial serum dilution was 1:8, not 1:100 as specified in the package protocol. IFA. The IFA test was performed in a standard manner (37). The AD-169 strain of CMV, cultured on L-645 human lung fibroblasts, was used. L-645 cells were originally obtained from the California State Department of Health Virus and Rickettsial Diseases Laboratory and are maintained in the Loma Linda University Virus Laboratory. Sera were tested in a serial twofold dilution pattern, from 1:8 through 1:256. Goat fluorescein-conjugated anti-human immunoglobulin G (lot no. 78077; Meloy Laboratories, Springfield, Va.) was used. A positive reaction was indicated by specific nuclear fluorescence observed by microscopic examination with UV light. The antibody level was expressed as the dilution or titer which maintained a 1+ fluorescence, as determined by the reader. Uninfected cells included in the cell spot preparation served as internal negative controls. IHA. The IHA procedure was basically performed as outlined by Yeager (40). This procedure differed from earlier IHA tests for CMV (4) in that glutaraldehyde-fixed human type O erythrocytes rather than sheep erythrocytes were used. This method had the advantage that glutaraldehydefixed, tanned, sensitized cells could be frozen for long periods until needed for the test. Freshly drawn whole human blood from a donor with type O erythrocytes was mixed with sodium citrate, washed several times in phosphate-buffered saline, and fixed in 1% glutaraldehyde (EM grade; Sigma Chemical Co., St. Louis, Mo.) for 30 min. After fixation, the cells were again washed and could be used immediately or stored frozen at -70°C for up to 6 months. The antigen used was CMV strain AD-169 grown in L-645 cells and prepared by sonic treatment of the infected cell suspension after one cycle of freezing and thawing. The optimal dilution of antigen was determined by titration against known positive sera, using cells treated with the optimal tannic acid concentration for each cell batch. The optimal antigen dilution was 1:8 in our preparation. The optimal tannic acid concentration was determined for each cell batch by testing known positive and negative sera against cells with tannic acid concentrations in the range of 1:20,000 to 1:320,000 (wt/vol) and sensitized with the optimal dilution of antigen. The optimal tannic acid concentration in our test system was 1:40,000. Once the correct antigen and tannic acid concentrations were determined, the fixed, tanned, sensitized human O erythrocytes and control cells (cells treated in like manner but not sensitized with CMV antigen) were stored frozen until needed for the test. The cells were added to serial twofold dilutions of serum in 0.15 M phosphate-buffered saline (pH 7.2) containing 1% normal rabbit serum and 0.1 M lysine. The serum dilutions ranged from 1:8 to 1:1,024. The cells were allowed to settle for 45 min at room temperature and then read. The titer was considered to be the highest dilution of serum that caused a 3+ to 4+ agglutination. RESULTS
positive by the other test procedures, and the remaining AC sample was shown to be negative. With the exception of these four samples, all sera that were positive in one test system were also positive in each of the other tests. Conversely, all sera giving negative results in one test gave the same result in each of the other methods. The results given by all tests on these sera were qualitatively the same. The titers obtained demonstrated the quantitative differences between the four methods. The CF test generally gave lower titers than the other procedures (Fig. 1). The data showed that each of these procedures offered a higher degree of sensitivity than did the CF test, as evidenced by the 1024
o
512
0
00
00
o
o
0
4
256
w
128
o
o
000
64
00
00
00
é-
000
0
00
32 16
8
